J. Phys. Chem. 1993,97, 11617-1 1624
11617
Infrared Matrix Isolation Studies of the Reactions of Dichloro- and Dibromomethane with Atomic Oxygen C. Lugez, A. Schriver, and L. Schriver-Mazzuoli' Luboratoire de Physique Mol8culaire et Applications, CNRS UPR 136, Universits Pierre et Marie Curie, Tour 13, Bte 76, 4 Place Jussieu 75252 Paris Cedex 05, France
E. Lasson and C. J. Nielsen Department of Chemistry, University of Oslo, P.O. Box 1033 Blindern. N-0315 Oslo, Norway Received: June 21, 1993; In Final Form: August 23, 1993'
The reactions of atomic oxygen with CH2C12 and CH2Br2 trapped in argon matrices have been studied by FTIR spectroscopy. O(lD) and O(3P) were generated in situ by UV photolysis of co-deposited ozone. Products were identified by employing l80and scrambled l80/l6Oozone as well as deuterated methylene halides. Kinetic studies performed on both CH2C12 and CHzBr2 with O(lD) under the same experimental conditions allowed the reaction pathways to be determined. With CHzBrz as parent molecule, three routes were evident leading to (i) CHOBr, (ii) CO-.(HBr)2, and (iii) CH2O. With CH2Cl2 as parent molecule, only the first two channels were observed. Carbonyl compounds rapidly decomposed under irradiation, and CO.-(HX)2 was also produced as a secondary species.
Introduction
The AirborneAntarctic Ozone Experiment, conducted in 1987, showed that depletion of polar ozone is caused by extremely high levels of C10 derived from industrial chlorofluorocarbonswith significant contributions from BrO derived from both natural and industrial sources.' Very high levels of C10 and enhanced concentrationsof BrO were also detected in the Arctic during the first Airborne Arctic Stratospheric Expedition (AASE-I) carried out in 1989.2 It is known that the ozone loss each spring over Antarctica is in large part caused by chlorine and brominecatalysis and also that the enhanced chlorine catalysis results from the alteration of the chemical environment near and inside the polar vortex by heterogeneous proce~ses.~ It is therefore important to obtain a detailed knowledge of all the different atmospheric degradation mechanisms of halogenated hydrocarbons. In this paper, we present results from infrared matrix isolation (IRMI) studies of the chemical reaction of dichloro- and dibromomethane with oxygen atoms. Some partial results for the CH2Br2 O3reaction have recently been presented: but the aim of the present paper is to compare the reactivity of CH2C12 and CH2Br2 under the same irradiation conditions. These compounds were mainly chosen as model compounds for more complex halocarbons, but they also occur themselves in the atmosphere. Dichloromethanehas a troposphericconcentration of approximately 35 pptv and a lifetime of half a year. The bromo compound has a tropospheric concentration of only 2-3 pptv.3 Even though this is a very low value, it is known that bromine in extremely low concentrations can enhance the ozone depletion by up to 2 0 % ~ . ~ The results will be discussed in comparison with those obtained in our previous studies concerning the reaction of oxygen atoms with CBrC13,6 CHC13,7 and CFC13.8
+
Experimental Section
The samples of CH2Cl2 and CH2Br2 were standard p.a. laboratoryreagents,while the deuterated samples originatedfrom Eurisotop and MSD, respectively. The samples were purified by distillation and checked by GC-MS before use. Ozone was Author to whom correspondence should be addressed. Aktract published in Advance ACS Absrracrs, October 1, 1993. 0022-3654/93/2097-11617$04.00/0
prepared from oxygen by electrical discharge in a closed system and trapped in liquid nitrogen. 180-enrichedsamples of ozone were prepared from 96 atom % '*OZ and 50 atom 76 1 8 0 2 originating from ICN Biomedicals and Eurisotop. Unfortunately, our ozone samples invariably contained traces of CO and C02. The matrix isolation spectra were recorded using two different experimental arrangements: The Paris system consisted of a Model 202A closed-cycle displex unit from Air Products equipped with a gold-plated mirror and a Bruker IFS 113v FTIR working in reflection mode. The Oslo System consisted of the same type of displex unit but was set up for transmission measurements employinga Bruker IFS 88. The samples were mixed with argon by standard nanometric methods and sprayed onto the 20 K cold mirror/window using dual-nozzle systems, and the spectra were subsequently recorded at 10 K with 0.5-cm-1 resolution (Paris) and at 15 K with 1-cm-1 resolution (Oslo). The deposited sample mixtures were irradiated with the UV light from medium-pressure mercury lamps. Some photolysis experimentswere performed without optical filters. However, in most of the experiments a 280-nm cut-off filter was inserted to prevent photolysisof the parent halide as well as possible secondary photolysis. With this filter, both O(3P) and O(1D) are formed from ozone, but experiments were also carried out employing a 360-nm filter whereby only O(3P) results. An experiment was performed with dichloromethane in solid oxygen (1/ 1000). The photolysis was performed with the full output of a mercury lamp. Irradiation of solid 0 2 with UV irradiation below 280 nm produces ozone and O(3P) followed by subsequent secondary photolysis of the ozone producing both states of atomic oxygen. Results
Vibrational Spectra of the Parent Molecules. The observed fundamental bands of matrix-isolatedCH2Cl2, CD2C12, CH2Br2, and CDzBr2 have been collected in Tables I and 11, which also contain data fromearlier vapor-phase9J0and matrix isolation'l-15 studies. After a short annealing of the matrices to 25 K, new bands appeared near the original bands, Le., for CH2C12: 3053 (~g),2992.4(~1), 1423.4-1421.2(~2),1268.7 (YE), 1160(~~),746.5748.5 (ug), and 712.5-709.4cm-1(~3). Thesame bands werealso 0 1993 American Chemical Society
11618 The Journal of Physical Chemistry, Vol. 97, No. 45, 1993
Lugez et al.
TABLE I: Com arison of Fundamental Vibration Frequencies (in cm-1) for Isotopic Methylene Chlorides in Vapor Phase and Amon Matrix. 8bserved and Calculated Frequency Shifts (in cm-I) vapor-phasedata" argon matrix data King this work 12CH235CI"CI A('ZCH235C137Cl) A('zCH237C137Cl)dAndrews Mikulec ~~
AI
VI
A2
~2 v3 v4 US
BI
V6
Bz
v7 vg vg
~
WD235CPCl AI
VI
vz v3 v4
A2 BI
v5 v6 v7
Bz
us v9
a
0.0 0.0 -6.0
0.0 0.0 -2.9 -3.4 -0.2 0.0 -0.3 -0.2 -2.2
2997.66 1435.0 712.9 281.9 1153b 3055b 898.66 1268.86 759.82
-6.8 -0.3 0.0 -0.6 -0.3 4.8
2998
1276 747
2993 1423c 710/706/704
2993.6 (0.2) 1428.4 (0.07) 7 12.91709.11706.6 71 1.6/708.7/706.1 (0.28/0.2/0.05)
3059 894 1265 74717461744 749.31747.31744.4
1156.8 (0.03) 3056 (0.001) 894.4 (0.1 1) 1266.2 (1) 749.3/746.0/744.0 (0.88/0.68/0.11)
A(12CD235C137Cl) A ( I ~ C D Z ~ ~ C P ~ CAndrews I)~ -0.0 -0.0 -6.4 -6.6 -0.5 0.0 -0.8 -0.1 -5.4
0.0 0.0 -3.1 -3.1 -0.2d 0.0 -0.4
2205.65 1060.81 687.6 279b 825 2303.72 713b 961.04 730.26
-0.ld
-2.9
this work
2205
2202.6 (0.09) 1055.6 (0.01) 685.31682.11679.4 (0.1/0.09/0.03) 826.9 (0.013) 2304.0 (0.001) 707.2/704.8 (0.013/0.011) 956.21955.1 (0.47/0.44) 72 117 19.417 16.6 (1 /0.98/0.17)
957 723
Reference 19. b Reference 30. Changed assignment. dCalculated (see text).
TABLE Ik Comparison of Fundamental Vibration Frequencies (in cm-l) for Isotopic Methylene Bromides in Vapor Phase and Arnon Matrix. Observed and Calculated Freuuencv Shifts (in cm-9 vawr-Dhasedata" argon ., matrix data 12CH279Br79Br A(12CHz79Br81Br)d A(12CH281Brg1Br)d Kelsall Andrews Young this work ~~
r
r
-0.6 -1.1
0.0 0.0 -0.1 -0.8 -0.5
A(12CD279Br79Br) 2214 1032 559 172b 790 2324 625 907 6Ogc a
0.0 0.0 -1.1 -2.1 -0.7 0.0 -0.1 -0.7 -1 .o
0.0 0.0
3009 1382 588 169b 109Y 3073 812 1195 653
A(L2CD279Br81Br)d 0.0 0.0 -0.6 -1 .o 0.0 0.0 -0.1 0.0 -0.5
301 1
3005.5 (0.006) 1400.1 (0.005) 583.0 (0.04)
582.971582.441581.92
649.411648.961648.48
A(1ZCD281Brs1Br)d 0.0 0.0 -1.2 -2.1 -0.1 0.0 -0.1 0.0 -0.7
1195 650
3074 811.5 1194.5 649.0
1096.8 (0.007) 3073 (0.007) 811.4 (0.05) 1194.2 (0.51) 648.9 (1 .O)
Andrews
this work
2208
2204.7 (0.012) 1034.1 (0.005) 555.4 (0.12)
905 62 1
781.5 (0.007) 2316.1 (0.015) 622.9 (0.22) 903.0 (1 .O) 618.6 (0.92)
Reference 10. Raman liq. Liquid phase. Calculated (see text).
observed by increasing the CHlClz concentration and have been attributed to molecular aggregates. In contrast to this, the IR spectra of co-deposited methylene halide and ozone showed no bands that could not be attributed to the parent molecules, and apparently there is no significant interaction between the two molecules. The infrared spectra of matrix-isolated ozone and l S 0 substituted ozone have been studied extensively (refs 16-18 and references therein). Our spectra were in good agreement with the literature. Vibrational Spectraafter Photolysis. Preliminary experiments where matrix-isolated dichloro- and dibromomethane were exposed to prolonged UV irradiation with X > 250 nm showed no detectable degradation of the samples. After irradiation of a CH~Cl2/03/Arsample (1/3/800 ratio) for 2 h by the unfiltered light from the UV source, approximately half of the ozone had vanished, and at the same time several new features appeared around 2800,2150, and 1750 cm-1 (Figure 1). The observed wavenumbers and optical densities are listed in Table 111,which alsocontains the data obtained from experiments with isotopic species. The bands at 28 11.2 and 278 1.1 cm-1 shifted
to 2035.4 and 2012.3 cm-1 with deuterium substitution. The three weak bands in the carbonyl region at 1756.7, 1754.6, and 175 1.3 cm-1 (the latter the strongest of the three) were shifted to 1715.4, 1713.2, and 1710.1 cm-l, respectively, when **03 was used as precursor and to 1731.4 and 1724.4 cm-1 when the deuterated dichloromethane was used. When scrambled lSO/ I60 ozone was used as precursor, the spectrum revealed the same bands as in the separate 1 6 0 3 and 1 8 0 3 experiments and with equal intensities. Temperature increase revealed different behavior of the triplet components in the carbonyl region as illustrated in Figure 2, wherea sequenceof temperatureand photolysiseffectsis presented in a CH2C12/l603experiment. The feature at 1756 cm-1 turned out to disappear quickly upon UV photolysis-even when using the 280-nm filter-but reappeared as the temperature was increased above 30 K. After annealing a t 40 K, it grew strongly accompanied by a new absorption at 1763 cm-1. The weak component at 1754.6 cm-1 appeared insensitive to temperature increase, whereas the main component at 175 1.3 cm-l and also the 1756.7-cm-1 absorption showed an intensity increase with temperature.
Reactions of Atomic 0 with CH2Cl2 and CH2Br2
The Journal of Physical Chemistry, Vol. 97, No. 45, 1993 11619
1766.0
I
I
1
1750 1725 tbvmulber cm-1
17020
Figurel. Detailed FTIRspectraofCHlCllandnaturalandisotopicOatommatrixreactionproductsinthe280&2750-,2170-2100-,and 1765-17OO-cm-~ regions: (a) after deposition at 20 K of CH2C1~/03/Ar= 1/1/1OOO mixture, (b) l s 0 3 after irradiation with the unfiltered medium-pressure mercury lamp for 1 h, and (c) l 6 l 8 0 3 (50%-50%) after irradiation with the filtered medium-pressure mercury lamp for 3 h. All spectra are recorded at 11 K.
1756.0
w
a 3d
B f
dd
ii d
B
d
Ei
F, 1702.5
1770
17m
1750
uawfu&R
1740
m 7
0 '
Figure2. Temperature effectson the CHzCl2 and 0 atom reaction products in the carbonyl region: (a) after depositionof CHzC12/03/Ar = 1/2/1000, (b) after irradiation with medium-pressure mercury lamp and cut-off filter WG 280 for 1 h with recording temperature T,= 11 K,(c) part b recording at 31 K, (d) part c recording at 12 K, (e) after second irradiation at 12 K with previous filter for 1 h and 30 min, and (f) part e after annealing at 40 K, recording at 11 K
TABLE III: Main Absorptions (cm-l) Formed after Medium-Pressure Mercury Lamp Irradiation of Argon Matrices Containing Natural or Isotopic Ozone and Hydrogenated, Deuterated Methylene Chloride CH2C12 + '603
CH2Cl2
+ "03
CDZClz + '60,
assignment V(HCl), (HCI)pCO
2811.2(0.101) 2781.1 (0.03)
2811.2 2781.1
2035.4 2012.3
2156.6 (0.08) 2155.6 (0.03) 1756.7 (0.004) 1754.6 (0.008) 1751.3 (0.04)
2105.3
2158.0
u(CO), (HC1)yCO
1715.4 1713.2 1710.1
1731.4 1724.4
u(C=O),
(HCICI=O.-HCl)
* In parentheses are peak height absorbance values for the CH2C12 + '603 reaction; irradiation time: 4 h). The analogous experiment with an argon matrix containing CH~Br2/03/Ar in a 1/2/1000 mixing ratio and photolyzed for 1 h revealed new bands around 2500,2150, and 1800-1700 cm-1 (Figure 3). The spectral features obtained from experiments with different isotopic species are summarized in Table IV. The
bands at 2499.3 and 2484.8 cm-1 shifted to 1794.9 and 1784.8 cm-I, respectively, when using the deuterated analogue. The four main bands in the carbonyl region appeared at 1799.5, 1756.3-1754.9, 1742.5, and 1727.7 cm-'. The last band was already present after deposition but seemed to grow with irradiation. Upon prolonged UV photolysis employing a 280-nm low-pass filter, additional weak bands were observed at 1781 and 1761 cm-I. After annealing the already photolyzed matrix, all these bands diminished or disappeared, and two bands at 1766 and 1759 cm-1 with numerous weaker features at 1750.2,1748.1, 1738.9, 1735.9, and 1719 cm-1 appeared. In the experiments with 1 8 0 3 , theprincipalfeaturesshifted to 1742.9,1715.4,1708.3, and 1695.8 cm-I, respectively, and when CDzBr2 was used as starting material, the bands were shifted to 1746.2,17 19.2,1696.1, and 1686 cm-1. Growth Behavior of the Products. A series of photolysis experiments without a filter were performed after several irradiation periods (i) to find primary and secondary products, (ii) to compare reactivity of bromine and chlorine compounds,
11620 The Journal of Physical Chemistry, Vol. 97, No. 45, 1993 0.5
Lugez et al.
I___..
-0.05
WAVENUMBER CM-1
WAVENUMBER CH-1
WAVENUMBER CM-1
Figure 3. Detailed FTIR spectra of CH2Br2 and '60 atom matrix reaction products in the 2520-2460-, 2180-2100-, and 1820-1720-~m-~regions: (a) after deposition of CH2Brl/Oa/Ar = 1 /2/800 mixture, (b) after irradiation with a medium-pressure mercury lamp filtered at 280 nm for 1 h, and (c) after irradiation as above for 8 h.
TABLE I V Main Absorptions (cm-') Formed after
Medium-Pressure Mercury Lamp Irradiation of Argon Matrices Containing Natural or Isotopic Ozone and Deuterated Methylene Bromide CH2Br2+ ' 6 0 3 CH2Br2+ 1*03CD2Br2 + I6O) 2499.3 (0.25) 2484.8 (0.13) 2153.1 (0.16) 1799.5 (0.02) 1772.8 (0.03) 1771.4 (0.003) 1756.3 (0.04) 1754.9 (0.03) 1742.5 (0.01) 1727.7 (0.01)
2499.1 2485.0 2140.3 1742.9
1794.9 1784.8 2154.1 1746.2
0 250
0 225 0 0 2200 00
assignment v(HBr), (HBr)Z-CO
0 175
v(CO), (HBr)2-CO v(C=O), HBrCO
0 125
16
2499cm-1
/
m
2 811cm-1
'2153cm-1
0 150 A
2157cm-1
no
loo
0 o 075 1715.4
1719.2
1708.3 1695.8
1696.1 1686.0
Y(cB~!:HBrCO r(C=O), H2CO v(C=O), HBPHZCO
In parentheses are peak height absorbance values for the CD2Br2 + 1603 reaction: irradiation time: 4 h.
0 050 0 025 0.025 0.000 0 000 0.0
and (iii) to examine the possibility of an isotopic effect. The photodissociation of ozone depends upon the position and power of the lamp. Hence, the growth of the species was measured relative to the ozone destruction rate. Furthermore, comparison of bromine and chlorine reactivity was made in the same experiment from a 1 :1 mixture of CHzCl2 and CH2Br2. In the same way, an accurate detection of a possible isotopic effect was performed using a 1:l mixture of CH2Clz and CD2C12. CH2C12 CHzBr2/O,/Ar. Figure 4 clearly illustrates that, for the two species, two different groups of bands can be extracted from the spectra. The first one appearing in the carbonyl region showed first an increase and subsequently a decrease in intensity, whereas the second one in the 2000-3000-~m-~region reached a nearly stationary state after the total depletion of ozone with, however, a weak decrease of the 2499- and 2153-cm-1 bands and a weak increase for the 281 1- and 2157-cm-1 absorptions. These two distinct behaviors suggest that the second group of product bands could arise from secondary chemistry initiated by photolysis of carbonyl products. Furthermore, experimental curves show that the concentration versus photolysis time behavior is slightly different for dichloromethane and dibromomethane. The results will be discussed quantitatively later. CH2CI2 + C&C12/03/Ar. Figure 5 presents the relative intensities of the characteristic product bands at 281 1.2, 1751.3 cm-*(parent moleculeCH~C1~) andat 2035.4,1724.4 cm-1 (parent molecule CD,Cl,) obtained after different irradiation times of a CH2C12/CD2C12/03/Ar (1 / 1/2/ 1OOO) mixture. If we take account of the fact that the absorption coefficient ratio for H / D stretching mode is about 2 but nearly the same for C 4 bands,
600.0
800.0 0
t(min)
a
0.035 0.030
0 025
+
t(min) Figure 4. Time evolution of the main products obtained from a mixture of CH2Cl2 and CH2Brz with 0 3 (CH2C12/CH~Br2/03/Ar = 1/1/2/ 1OOO) under irradiation with an unfiltered medium-prmsure mercury lamp: (m) experimental points relativeto CH2Clz as parent molecule and (A) experimental points relative to CH2Br2 as parent molecule.
the observed results suggest an equal abundance of the carbonyl species and a similar photolysis behavior when both CH2C12 and CDzClz are used as parent molecules. Reaction with Ground-State Atomic Oxygen. In order to determinewhether the differentobserved reaction products arrived with both 3Pand IDatomic oxygen, an experiment was performed where a mixture CH2C12/CH2Br2/03/Ar in the ratio 1/1/2/ 1000was irradiated for 8 h with light from a Xenon lamp filtered by a 360-nm cut-off filter. Figure 6 compares the spectrum with previous spectra obtained when CH2Cl2 + 09 and CHZBr2 0,
+
Reactions of Atomic 0 with CHlCl2 and CH2Br2
f
The Journal of Physical Chemistry, Vol. 97, No. 45, 1993 11621
@
d
7 N
'1762.0
1750
1725
1710.0
Uaverimbcr cn-'
Figure 5. Comparison of the evolution of the products with the time of irradiation (unfiltered medium-pressuremercury lamp) in 280CL1980-cm-1 and 176&1710-~m-~ regions after deuteration of the parent molecule: CH2C12/CD2C12 = 1, CH2C12,CD~C12/03/Ar= 1/2/1000. Irradiation times: (a) 0, (b) 1 h, (c) 2 h, (d) 3 h, 40 min, (e) 5 h. According to the parent molecules, products are labeled either H or D.
0d
i3 H
d
m l o o a I 7 a m 8 2 1 0 0 2 a I s o a a s l a x r z s z s z s s D = z s D o m I
I
I
I
1
I
L
9
I
1
I
1
I
1
-
6,
1;: d
8, d
8, ana
d
as a a as am ars
-
a3b3 ims 1790 1m 1710 1180 1m 1 m o umndr a ' m' Figure 6. Comparison of the spectra in the key regions after irradiation of (a) CH2C12/03/Ar (1/1/1000) mixture with medium-pressuremercury
lamp without filter, (b) CHzBr2/03/Ar (1 /1/ 1OOO) mixture with medium-pressure mercury lamp without filter, (c) CH2Clz/CH2Br2/03/Ar (0.5/ 0.5/1/1000) mixture with Xenon lamp and cut-off filter WG 360. Irradiationtimes: (a) 40 min, (b) 10 min, (c) 7 h, 30 min. All spectra are recorded at 11 K. matrix mixtures were photolyzed with the full light of the mediumpressure mercury lamp. Two observations deserve attention: (i) in the carbonyl region the band a t 1742 cm-I was not observed and (ii) in the 2800-2150-cm-1 region bands previously reported appeared as traces. Oxygen Matrix. In an attempt to gain information on the reaction intermediates, an experiment on the photolysis of CH2Clz in a solid 0 2 matrix, which is a strongly oxidizing environment, was carried out with the full light of the mercury lamp (CH2C12/02 1/1000). As mentioned in the Experimental Section, irradiation of solid 02 below 280 nm produces ozone and O(3P)
with subsequent secondary photolysis of ozone. Although the product yield was considerably lower than that in the previous experiments, similar product absorptions (compared to argon matrix experiments) appeared in the three key regions after photolysis. They were respectively observed a t 2804.6, 2771.3; a t 2159.6; and a t 1754.4, 1753.0, 1749.5 cm-I. The triplet in the carbonyl region was sensitive to temperature increase but with a trend different from that observed in the argon matrix: the lower component grew with the temperature whereas the higher component disappeared. In addition, new bands not seen in argon were observed a t 1724.4, 1391.3, and 1100.0 cm-I.
Lugez et al.
11622 The Journal of Physical Chemistry, Vol. 97, No. 45, 1993
Mscussion Vibrational Spectra of Parent Molecules. The bands observed for the parent halides are in agreement with values from the vapor-phase9J0and matrix isolation studies.l2J4J5 Some of the observed bands showed isotopic splitting due to the natural abundance of chlorine and bromine isotopes (3SCIWI = 57% W137C1= 37%, 37C137C1- 6%, 79Br79Br 26%, 81BralBr 24%, and 79Br8lBr 50%). This have previously been described for the v3 and u9 bands.9JlJ2 For comparison, we have included the calculated isotopic shifts in Tables I and I1 (see later). Due to the isotopic mixing, the us mode (A2, IR inactive) is also observed in the spectra. In a recent report of the IRMI spectrum of CH2C12,13the v2 mode was assigned to a band at 1455 cm-I while the 1428-cm-1band was ascribed to the overtone, 2v3. Because of the better resolution in our experiments, we observed an isotopic splitting of the 1450-cm-1 band (1453.5,1451.5, and 1448.7cm-l) while the 1428.4-cm-1 band shows no such splitting. We therefore assign the latter band to the v2 mode (CH2 scissoring), for which the normal coordinate calculations show no isotopic splitting. The bands around 1450cm-I are instead ascribed to a combination of the two CCl stretching modes, v3 v9 (B2). This interpretation is also in agreement with the thorough vapor-phase study by Duncan et ale9 The analogous combination band in CH2Br2 is observed at 1226 cm-I, also in good agreement with the literature, where this is reported as the strongest combination band.10 The vibrational wavenumbers for methylene chloride as well as the isotopic shifts and a general harmonic force field have been presented by Duncan et al.19 In order to calculate the vibrational fundamentalsand isotopic shiftsin methylene bromide, we adopted the electron diffraction structure20 and scaled the methylene chloride force field by the inverse square of the distances for the stretching constants and the stretch-stretch interactions and by the inverse distances for the bending constants and the stretchbend interactions. Identification of Compounds Formed by UV Photolysis. Identification and vibrational assignments of the two product sets recognized from their growth behavior (Figure 4) will be considered separately. The first one was characterized by very weak, photosensitive bands in the carbonyl region, whereas the second one appeared in the 2800-2000-~m-~region as stronger absorptions. Carbonyl Products. After photolysis of the CH2C12/03 mixture, a triplet band was observed at 1751, 1753, and 1756 cm-1 with a 42-cm-1 1 8 0 shift and a 27-cm-1 deuterium shift. This group of bands has been interpreted as due to the C=O stretching vibration in formyl chloride complexed by HCI. This agrees with previous identification of the complex in solid nitrogen by Nelander.21 The multiplicity of the C = O band also observed in nitrogen suggests a possible different relative orientation of the complex forming molecules. The increasein intensity of the bands with the temperature is surprising. It is perhaps an indication of recombination in the matrix cage of photolytic fragments not identified. Photolysis of the CH2Br2/03 matrix gave three new bands in the carbonyl stretching region at 1799, 1756-1754, and 1742 cm-I. As can be seen in Figure 4, their different growth behavior indicates that they belong to three distinct species. The band at 1742 cm-I shifts to 1696 cm-I upon deuteration and is due to UCIO of free formaldehyde.22The position of the doublet at 17561754 cm-I and its isotopic shift upon deuteration (1719 cm-1) resemble the situation in the CHOCl-HCl complex and have tentatively been assigned to the CHOBr-.HBr complex. The identification of the 1799-cm-1 band is less straightforward. However, the apparent isotopic shifts observed with 1 8 0 3 and under deuteration indicate the presence of one hydrogen atom and one oxygen atom, and the absorption is therefore tentatively ascribed to free HCOBr. Formyl bromide has not been matrix isolated previously, but the IR spectrum of it in the vapor phase
-
-
+
-
was recently obtained by Niki.23 The carbonyl vibration of formyl bromide was observed at 1799.5 cm-l in Fermi resonance with 2u6at 1774.2cm-I. Itshiftsto 1746.2cm-IintheDCOBrspecies. The product yield in our experiments is probably too small for an accurateobservation of the 2v6band, although a narrow feature at 1771.3 cm-I which grows in concert with the 1799-cm-1 band could be due to this overtone. Furthermore, the shift observed for isotopic substitution with 'EO3is larger than the one observed for the HCOBr-HBr complex, but it is of the same order of magnitude if it is calculated from the 1789-cm-1 position, the position for the fundamental carbonyl band vibration in the vapor phase. Second Set of Products. With dichloromethane as the parent molecule, new bands appeared at 281 1 and 2781 cm-1. Upon deuteration, these band shifted to 2035 and 2012 cm-1 and can unambiguously be assigned to the HCI stretching vibration of the (HC1)2-.C0 aggregate.24s25The corresponding CO absorption was observed at 2156.6 and 2155.6 cm-I for the parent and deuterated molecules, respectively. The results for dibromomethane have been interpreted in the same fashion. The new bands at 2499,2485 (shoulder at 2479 cm-I), and 2153-2152 cm-I for the parent molecule, shifting to 1795, 1785, and 2154 cm-1 with the deuterated bromide and to 2499,2485, and 2 140cm-l with ls03 precursor, have been assigned to the (HBr)2-.CO complex. Perchard et observed the same bands in a study of various HBr/CO complexes, but their interpretation was somewhat different. In experimentswith CD2Br2/O3, the DBr stretching modes of the ( D B r ) p C O complex fall in the carbonyl region (1795 and 1785 cm-I). Growth curves show them as final products, however, and they are therefore easily distinguished from the carbonyl intermediates. Kinetic Studies. In order to obtain quantitatively significant results for the abundance of the different products as a function of the time and to determine accurately primary and secondary species, kinetic studies were performed using the experimental curves of Figure 4 within the framework used in our previous ~ t u d i e s .Reactions ~ are considered as first-order processes and can be schematically represented by
ko O3-O2+0
-ki
k2
(CH,X2, 0,) P
S
where (CH2X2, 0 3 ) refers to the fraction No of CHzXz and 0 3 in neighboring "reactive sites", P to primary products, and S to secondary products. The rateconstant ko of ozonedepletion can be easily calculated from (0,)' = ( 0 3 ) 0 exp(-kot). The abundances of P and S are respectively given by NOkl Np= -[exp(-k,t) - exp(-k,t)] =
kl - k2
N,=NO[l+
k2 exp(-k,t) - k, exp(-k,t)
kl - k2 Numerical values of a,kl, and k2 for carbonyl products were obtained by a least-squares fit of experimental curves using the Microsoft Excel program. Values were adjusted in order to minimize the residual sum of squares between the calculated Np and observed Npvalues. Results are summarized in Table V and compared to the average dissociation rate constant ko of ozone. kl values are of the same order of magnitude for the formation of all carbonyl compounds, suggesting that the reaction is principally governed by ozone photodissociation. Secondary photolysis of COHCI.-HCl appears slightlymore rapid than that
The Journal of Physical Chemistry, Vol. 97, No. 45, 1993 11623
Reactions of Atomic 0 with CHzClz and CHzBrz
TABLE V Comparison between Kinetic Rate Constants (in min-l) at 10 K for Formation ( k l ) and Dissociation (k2) of the Primary Products Obtained from Photolysis of a CHzCIJ CHzBr2/0JAr (1/1/2/1000) Mixture by Medium-Pressure Mercurv Lamp parent molecule products ki k2 kl/kQ CHiCli .H 2.9 x 10-2 7 x 10-3 1.2 - -
TABLE VI: Predicted A I P for CH2X2 + 0 Reactions (X = Cl, Br) in kJ/mol
x = CI reaction CHzXz + O(lD) CHzO + Xz CHzXz + O(1D) CHOX + HX CHzXz + O(lD) COX2 + Hz CH~X + o(3~) ~ = C H ~ O+ xz CHzXz + O(3P) = CHOX + HX CH~X + o(3~) ~ = coxz+ H~
CH...0-c‘
‘a CO*..(HC1)2 CH2Br2
2.5 X
10-’
1 7.5
L O X 10-2
...
%H O - C ,
HH
x 1 ~ 3 1.2
Br
HH
3.3 X le2 3.0 X
1W2
1.3
O=C\Br
CO-.(HBr)2
2.5 X le2
1
In this experiment,the average depletion rate constant of ozone, ko, was found qual to 2.5 X 1 t 2 min-I.
STO-3G 3-21G -247 -358 -330 127 16 44
-280 -420 -300 51 -89 32
200
500
t(min)
Figure 7. Comparison between calculatedcurvesfrom differentequations for the time evolution of of the CD-(HCl)2 product: (-) calculated curve using q 2 with kl = 2.9 X 10-2m i d and k l = 7.0 X lW3 m i d , (- - -) calculated curve using eq 3, and (*)experimental points.
of COHBr-SHBr, whereas the dissociation of free COHBr appears nearly as fast as its formation. Photolysis of these carbonyl compoundsis assumed to produce secondary products. In particular, COHX-.HX can lead to CO--(HX)z as final products and experimental growth curves of these species are expected to be fitted by eq 2 using the previously calculated kl and kz values for each parent molecule. As illustrated in Figure 7, the experimental growth curve for CO--(HCl)Z complex formation does not comply with that for CHOCl--HCl formation (kl = 2.9 X 10-2 min-’ and k2 = 7.0 X lO-3min-I). In fact, the experimental curve can be approximated by a single-exponential curve of the form Ns= No( 1 - exp(-klt)).
CHART I
HClCO+HCl
CO+2HC1
STO-3G
-278 430 -333 53 -99 -2
-320 403 -347 54 -29 27
The same applies to the CO.-(HBr)Z growth curve, with a similar rate constant of ca. 2.5 X min-l (kl/ko = 1). This suggests that CO.-(HX)2 complexes are formed both as primary and secondary products of the CHzXz + Os reaction. Photolytic Process. Reaction Pathways. From kinetic studies, the observed product distribution points to two pathways for the initial step of the CH2Cl2 + Os reaction and three channels for that of the CHzBr2 + O3 reaction (Chart I). In the matrix cage, the HX molecule appears complexed either by the formyl-halogenated or CO species, although in the case of the CH2Brz parent molecule free formyl bromide also seems to be present. Upon further photolysis, HXCO--HX complexes decompose to (HX)z-.CO species whereas H2CO dissociates to CO and H2, as expected.20 Only the first pathway leading to XHCO molecules is observed with atomic oxygen both in the excited and ground state. Ground-state atomic oxygen is not reactive enough to produce HzCO or CO-.(HX)z. Table VI summarizes ab initio calculated W values for different possible routes.26 Although there is some dispersion among the results obtained with the different basis sets, it is clear that CHOX, COXz, CO, and CHzO all would result from the reactions with O(1D) and also that CHOX formation should be a possibleproduct from a reaction with O(3P). COX2 is not observed in our experiments, and CHzO is only observed with CH2Br2 as the parent molecule, this last observation being the most striking result incomparing the CH2C12/03 and the CHzBrz/03 reactions. Before excluding this reaction path for the methylene chloride reaction, the possibility that H2CO-.HCl is formed and then rapidly destroyed needs to be examined. According to N e l a ~ ~ d e r , ~ ~ when nitrogen matrices containing formaldehyde and chlorine were photolyzed through a NiS04 solution, the formaldehyde complex decreased rapidly in intensity, and formation of HClCO was observed. Although this possibility cannot entirely be ruled out in our experiments, it seems unlikely for two reasons: (i) no traces of O=CHz or O-CHz-.HCl species were observed in our spectra and (ii) the growth curve of HC1CO:HCl has an exponential behavior without a sigmoidal contour at short times, suggesting that it cannot also be a secondary species produced by photodissociation of a HZCO-Clz complex. Mechanism. The primary step of the formyl halide formation might be one of the three following possibilities: (i) insertion of an oxygen atom into the C-H bond, (ii) abstraction followed by bond cleavage leading to the product, and (iii) addition toa carbon atom giving a five-centered excited intermediate. The insertion of an oxygen atom in an excited state into the C-H bond has already been observed; in particular, by photolyzing
+
O O
X = Br
3-21G*
HBrCO+HBr
3
C0+2HBr I
\
I
CO..
. (H%r)
H2CO+Br2
I CO+H2
11624 The Journal of Physical Chemistry, Vol. 97, No. 45, 1993
TABLE VII: Comparison of Reaction Products for Reactions between Various Halomethanes and Either O(lD) or OPPP) Oxvnen Atoms Drimarv maducts
CBrCl3 + 0
0 =c
o=c
CFClp + 0
0 =c
CC4 + 0
0 =c
/F
HU \U
\CI
-
\U
HU
a
Lugez et al. and O(ID). In all cases, carbonyl compounds are the main primary products. In the absence of a hydrogen atom in the parent molecule, two halogen atoms are abstracted. As expected, the C-Br (ca. 210 kJ/mol) bond is more easily broken than the C-CI bond (ca. 370 W/mol) while the C-F bond (ca. 530 kJ/mol) is not broken at all. Irradiation of a CC14/03/Ar or CBrd/Oa/Ar mixture leads to the expected product COX2. For the halogenated hydrogen species, the major route observed for the reaction with both O(3P) and O(lD) is the abstraction of H X species. This abstraction leads to H H X from CHzXz and to H X 2 from CHX3. Thus, the great affinity between the H and X atoms favors this abstraction of HX occurring also with O(3P).
\U
CBra + 0
0 =c
/Br
Acknowledgment. Our sincere thanks to Dr. A. J. Barnes for reading and improving the manuscript. Oneof us (L.S.M.) thanks the NAVF (Norway) for an invitation to the Department of Chemistry of Oslo University (Professor P. Klaboe). In Paris, our sincere thanks to Dr. R.Levant for helping in the calculations performed with the Microsoft Excel program,
a
\Elf
CHCla + 0
o=c \a
References and Notes CH2C12 + 0
0 =c
/H
o=c
/
H
‘U
\U,
O=C\Br
CHBr3 + 0
o=c
/H \E,
CH2Br2 + 0
CH3Br + 0
o=c
/H \Br
0 =c
/H \H
/H
o=c\H
Experiment not performed.
an 03-CH4 mixture in liquid argon the excited CH3OH intermediate was evident.27 However, this insertion is unlikely when atomic oxygen is in its ground state.** In our experiments, both O(3P) and O(lD) react in the same way, and insertion into the C H bond seems to be ruled out. The absence of any isolated intermediate makes it difficult to determine which of the two other pathways is the major one, although the two bands observed in the oxygen matrix (1391.3 and 1100.0 cm-l) assigned to the 02H species29 could be an indication of H abstraction followed by bond cleavage leading to the product. Formation of CH20 observed only in the reaction between CH2Br2 and O(lD) probably arises from a different mechanism. In this case, insertion of O(lD) into a C-Br or a C-H bond to form a vibrationally excited intermediate which undergoes two rearrangements before its fast dissociation is a possibility. Table VI1 shows a summary of the different products found in our laboratories for the reaction between halocarbons and O(3P)
(1) The Airborne Antarctic Ozone Experiment. J. Geophys. Res. 1989, 94, (D9, 30 August; D14, 30 November).
(2) The first Airborne Arctic Stratospheric Exedition (AASE-I). Geophys. Res. Lett. 1990, 17 (Number 4, March). (3) WMO, Global ozone research and monitoring project. Report. no. 25, 1991. (4) Lasson, E.;Klaboe, P.; Nielsen, C. J. Air Pollution Res. Rep. 1991, 42, 79. (5) Wayne, R. P. Chemistry of the Atmospheres, 2nd ed.; Oxford Scientific Publications, 1991. (6) Abdelaoui, 0.;Schriver, L.; Schriver, A. J . Mol. Srruct. 1992,268, 335. (7) Schriver, L.; Gauthier-Roy, B.; C a d r e , D.; Schriver, A.; AbouafMarguin, L. Chem. Phys. 1992, 163, 357. (8) Schriver, L.; Abdelaoui, 0.;Schriver, A. J . Phys. Chem. 1992, 96, 8069. (9) Duncan, J. L.; Nivellini, G.D.; Tullini, F. J . Mol. Specrrosc. 1986, 118, 145. (10) Dennen, R. S.;Piotrowski, E. A.; Cleveland, F. F. J . Chem. Phys. 1968, 49, 4385. (11) King, S. T. J. Chem. Phys. 1968,49, 1321. (12) Kelsall, B. J.; Andrews, L. J . Mol. Spectrosc. 1983, 97, 362. (13) Mikulec, J.; Cerny, C. Spectrochim. Acta 1987, 43A, 849. (14) Andrews, L.; Prochaska, F. T.; Auk, B. S . J. Am. Chem. Soc. 1979, 101,9.
(15) Young, N. A.; Speiccr, M. D. J . Mol. Srrucr. 1990, 222, 77. (16) Spoliti, M.; Nunziante-Cesaro, S.; Mariti, B. J . Chem. Phys. 1973, 59, 985. (17) Andrews, L.; Spiker, R. C., Jr. J . Phys. Chem. 1972, 76, 3208. (18) Green, D. W.; Erwin, K. M. J. Mol. Spectrosc. 1981, 88, 51. (19) Duncan, J. L.; Lawie, D. A.; Nivellini, G. D.; Tullini, F.; Fergusson, A. M.; Harper, J.; Tonge, K. H. J. Mol. Spectrosc. 1987, 121, 294. (20) Beagley, B.; Brown, D. P.; Freeman, J. M. J . Mol. Srrucr. 1974,20, 315. (21) Strandman-Long, L.; Nelander, B.; Nord, L. J . Mol. Srruct. 1984, 117, 217. (22) Diem, M.; Lee, E. K. C. J . Phys. Chem. 1982,86, 4507. (23) Yarwood, G.; Niki, H.; Maker, P. D. J . Phys. Chem. 1991,95,4773. (24) Perchard, J. P.; Cipriani, J.; Silvi, B.; Maillard, D. J . Mol. Srruct. 1983, 100, 317. (25) Andrews, L.; Arlinghaus, R. T.; Johnson, G.L. J . Chem. Phys. 1983, 78, 6348. (26) Trucks, G.W.; Foresman, J. B.; Schlegel, H. B.; Raghavachari, K.; Robb, M.;Binckley, J. S.; Gonzalez, C.; Defrccs, D. J.; Fox, D. J.; Whiteside,
R.A.;Seeger,R.;Melius,C.F.;Baker,J.;Martin,R.L.;Kahn,L.R.;Stewart, J. J. P.; Topiol, S.; Pople, J. A. Guussiun 90, Revision I; Gaussian Inc.: Pittsburgh PA, 1990. (27) Demore, W. B.;Raper, 0. F. J . Chem. Phys. 1967, 16, 2500. (28) Luntz, A. C. J . Chem. Phys. 1980, 73, 1143. (29) Milligan, D. E.; Jacox, M. E. J . Chem. Phys. 1963.38.2627. (30) Escribano, R.; Orza, J. M.; Montero, S.; Domingo, C. Mol. Phys. 1979, 37, 361.
